[postgis-users] loading mostly identical states-based shapefiles

[Puneet Kishor]

On May 26, 2012, at 8:56 AM, Nicolas Ribot <nicolas.ribot at gmail.com> wrote:

> Hi Puneet,
> 
> Concerning the first point, you will be able to control the column you
> want to load with the GDAL/OGR suite (gdal.org).
> ogr2ogr allows changing the structure of datasets during conversion.

This is handy information. For now, I wrote a Perl script that loaded each file into its own table, copied the common columns only to a master table, then dropped the individual tables. Was all done in a couple of minutes. 

Although, would be handy to have a switch to either include "named columns" or exclude "named columns in the shp2pgsql utility. That would make it analogous to the -n or -N switches in pg_dump/pg_restore.


> 
> The -S switch is very handy when you are sure the input shapefile
> contains only simple objects (points, linestrings, polygons) and you
> don't want your geom column to be a MULTI type:
> The shapefile format specifications allow to store both simple and
> multi objects in the same shapefile. Shp2pgsql generates multi objects
> layers by default, as it does not scan the shapefile to see if
> multiobjects exist, prior to loading it.


Thanks again for this great info, but this still begs the question: why would I want MULTI over SIMPLE features or vice versa? Performance, integrity, querying, what?

> 
> Nicolas
> 
> On 26 May 2012 04:29, Mr. Puneet Kishor <punk.kish at gmail.com> wrote:
>> I have US states shape files that are mostly identical, except for those silly legacy columns that have the state code prefix. Is there a way I can import these shapefiles via `shp2pgsql` specifying to *not* load a particular column? As is, the loader croaks because the '-c' switch creates the table on the first state load creating a column starting with that state's code, and then the second state can't find that column.
>> 
>> An additional question -- the '-S' switch promises to "Generate simple geometries instead of MULTI geometries." Why would I want to do that? Please educate me.
>> 
>> 
>> 
>> --
>> Puneet Kishor
>> _______________________________________________
>> postgis-users mailing list
>> postgis-users at postgis.refractions.net
>> http://postgis.refractions.net/mailman/listinfo/postgis-users
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users